Taken MCP Server

taken?

taken? answers one question before you volunteer for a GitHub issue: is it already taken?

In plain terms, it checks an issue before you start work: is it still open to contributions, is anyone already on it, and is the repo responsive enough that your PR will get reviewed.

Try it without installing anything: the live console runs the real checks in your browser.

The catch it was built for: GitHub shows "linked a pull request" events in the issue timeline, but never in the comments. You can read every comment on an issue and still miss that someone already opened a PR for it. taken? checks the timeline, scans comments for people claiming the work, looks at assignees, reads CONTRIBUTING.md for AI contribution policies, and checks whether the repo is still active. Then it gives a verdict, GO, TAKEN, or CAUTION, with the evidence cited.

Requirements

• Python 3.10 or newer
• The GitHub CLI (gh), authenticated

taken? only makes read-only API calls through your own gh login. It never sees or stores tokens, and it never writes anything to GitHub.

Install as a tool

uv tool install taken-gh
taken owner/repo#123

Or with pipx:

pipx install taken-gh
taken owner/repo#123

Usage

uv run taken owner/repo#123

A full issue URL works too:

uv run taken https://github.com/owner/repo/issues/123

Useful flags:

• --json: print the full findings as JSON instead of the human summary
• --me LOGIN: ignore your own comments when scanning for claimants
• --file PATH: read targets from a file, one per line
• --limit N: scan mode checks at most N open issues per repo (default: 20)
• --label LABEL: scan mode only considers open issues carrying this label
• --no-cache: bypass the API response cache
• --clear-cache: delete the API response cache and exit
• --graphql: force the GraphQL fetch path even when not logged in to GitHub (falls back to REST on failure); GraphQL is already the default for logged-in users (or TAKEN_GRAPHQL=1)
• --rest: force the REST fetch path even when logged in; escape hatch for the GraphQL default (or TAKEN_REST=1)
• --persistent-session: run the GraphQL query over one persistent HTTPS connection for the whole process; the token comes from gh auth token and is held in memory only. Opt-in (or TAKEN_PERSISTENT_SESSION=1); see GRAPHQL_NOTES.md for the security tradeoff
• --version, --help

Exit codes: 0 means GO, 1 means TAKEN, 2 means CAUTION, 3 means something broke (bad target, no gh, API error). With several targets the exit code is 0 when every target produced a verdict and 3 when any target failed.

Batch mode and repo scans

Give taken several targets and it prints one verdict line per target:

taken owner/repo#123 owner/repo#124 --me mylogin
TAKEN   owner/repo#123  assigned to: dk5488
GO      owner/repo#124  no linked PRs, no assignees, no claimants, repo is active

A bare owner/repo scans the repo automatically: its open issues (most recently updated first, PRs excluded) are each checked:

taken django/django --label "good first issue" --limit 10

API responses are cached for one hour in ~/.cache/taken (override with TAKEN_CACHE_DIR), so repeated scans stay cheap. --no-cache skips the cache; --clear-cache deletes it.

Discover mode

taken --discover finds contribution candidates across GitHub. It piggybacks on GitHub's issue search API (the same source the web aggregators use) for raw candidates, then runs taken's full verification on each one and ranks the survivors. Only GO verdicts make the list.

taken --discover --language python --min-contributors 3 --limit 10
  6  octocat/hello-world#42  maintainer replied; updated 1d ago; repo pushed 0d ago
  4  octocat/other-repo#7    updated 3d ago; repo pushed 2d ago

Candidates are scored on the signal no aggregator filters on: maintainer responsiveness. A non-author, non-bot comment scores +3; an issue updated in the last 7 days scores +2; a repo pushed in the last 7 days scores +1. Each line explains its own score. --label restricts the search to one label instead of the default set (good first issue, good-first-issue, beginner friendly, help wanted).

Every candidate is also marked for first-time friendliness: the issue's own labels (e.g. [good first issue]) plus repo-level signs that contributions are welcome ([has CONTRIBUTING.md], recent merged PRs). Repo scans annotate the GO recommendations the same way, friendliest first:

taken octocat/hello-world
  GO      octocat/hello-world#42  no claimant found
  ...
  1 GO candidate: octocat/hello-world#42 (good first issue)

Candidates are verified in parallel (8 workers by default; --jobs N tunes it). A progress bar on stderr shows live feedback during the run; --no-progress hides it. The bar never touches stdout, so --json stays script-friendly.

JSON output

taken --json owner/repo#123 prints an object with three keys:

• verdict: one of GO, TAKEN, CAUTION
• reasons: list of human-readable strings explaining the verdict
• findings: the raw check results:
  • target: e.g. owner/repo#123
  • issue: number, state, title, labels, assignees, comment_count, author, url, created_at
  • linked_prs: list of number, title, state, merged, author, url
  • claimants: list of author, date, pattern, snippet, url
  • ai_policy: verdict (ban, disclosure-required, or none-found), snippet, source
  • repo_health: pushed_at, pushed_recently, recent_merges, stars

MCP server

taken also ships as an MCP server, so coding agents can check issues with a tool call instead of shelling out to the CLI. It needs the same setup: gh installed and authenticated, and it only makes read-only API calls through your own login.

Run it directly (the MCP server ships with every install):

uvx --from "taken-gh[mcp]" taken-mcp

Or add it to your MCP client config:

{
  "mcpServers": {
    "taken": {
      "command": "uvx",
      "args": ["--from", "taken-gh[mcp]", "taken-mcp"]
    }
  }
}

Three tools:

• check_issue(owner, repo, issue_number, me?): GO/TAKEN/CAUTION verdict with reasons and full findings for one issue.
• scan_repo(owner, repo, limit?, label?, me?): the repo's open issues with verdicts, GO first, plus recommendations (just the GO targets), a verdict summary, and per-issue friendly_labels / welcoming markers so the safest issues to adopt stand out.
• discover_candidates(limit?, language?, label?, min_contributors?, me?): good-first-issue style candidates, verified and ranked, each marked with its first-time-friendly labels and contribution-welcome signals.

scan_repo and discover_candidates include effective_parameters in every response so clients can see the resolved defaults and filters behind the results. Their MCP input schemas also describe each optional parameter's default.

taken is published in the official MCP Registry as io.github.RogueAlg0/taken.

Examples

An issue with a PR already fixing it:

$ uv run taken Exodus-Privacy/exodus#300
taken? Exodus-Privacy/exodus#300
verdict: TAKEN
  ...
  why:
    - open PR #700 already covers this: https://github.com/Exodus-Privacy/exodus/pull/700

An issue someone is already assigned to:

$ uv run taken chahe-dridi/vscode-agent-bell#200
taken? chahe-dridi/vscode-agent-bell#200
verdict: TAKEN
  ...
  why:
    - assigned to: dk5488

A clean issue, with your own volunteering comment filtered out:

$ uv run taken snowflakedb/snowflake-cli#3145 --me RogueAlg0
taken? snowflakedb/snowflake-cli#3145
verdict: GO
  ...
  why:
    - no linked PRs, no assignees, no claimants, repo is active

How the verdict works

TAKEN wins over CAUTION, which wins over GO.

• TAKEN: the issue is closed, an open PR links to it, or someone is assigned.
• CAUTION: a comment says someone wants it (soft claim, they may have moved on), a linked PR was merged but the issue is still open, the repo has an AI policy to respect, the repo looks inactive, or a beginner-friendly label sits alongside a long discussion thread or a design-level label (worth skimming before you start).
• GO: none of the above.

The claimant scan is a heuristic over comment text, not proof. The verdict always prints its evidence so you can judge for yourself.

Development

uv sync          # install dev tools (ruff, pytest)
uv run pytest
uv run ruff check
uv run ruff format --check

Hotspots

The hotspots badge tracks files that are both complex and frequently changed: complexity is cyclomatic complexity summed per file (via radon); churn is the number of commits touching the file. A file needs attention when it has complexity >= 50 and >= 5 commits; these are the files where refactoring pays off most.

To regenerate the badge data locally:

uv run --python 3.12 --with radon scripts/hotspots.py

AI assistance

This project is built with AI assistance, and says so openly. Every contribution is reviewed and understood by its author before it lands.

Security

See SECURITY.md for the supported versions and how to report a vulnerability. Please use private vulnerability reporting, not a public issue.